Decreased Rubisco activity during water stress is not induced by decreased relative water content but related to conditions of low stomatal conductance and chloroplast CO2 concentration

Authors

  • J. Flexas,

    Search for more papers by this author
    • *

      Authors who contributed equally to this work.

  • M. Ribas-Carbó,

    Search for more papers by this author
    • *

      Authors who contributed equally to this work.

  • J. Bota,

    1. Laboratori de Fisiologia Vegetal, Grup de Biologia de les plantes en condicions mediterrànies, Universitat de les Illes Balears, Carretera de Valldemossa Km 7.5, 07122 Palma de Mallorca, Balears, Spain
    Search for more papers by this author
  • J. Galmés,

    1. Laboratori de Fisiologia Vegetal, Grup de Biologia de les plantes en condicions mediterrànies, Universitat de les Illes Balears, Carretera de Valldemossa Km 7.5, 07122 Palma de Mallorca, Balears, Spain
    Search for more papers by this author
  • M. Henkle,

    1. Laboratori de Fisiologia Vegetal, Grup de Biologia de les plantes en condicions mediterrànies, Universitat de les Illes Balears, Carretera de Valldemossa Km 7.5, 07122 Palma de Mallorca, Balears, Spain
    Search for more papers by this author
  • S. Martínez-Cañellas,

    1. Laboratori de Fisiologia Vegetal, Grup de Biologia de les plantes en condicions mediterrànies, Universitat de les Illes Balears, Carretera de Valldemossa Km 7.5, 07122 Palma de Mallorca, Balears, Spain
    Search for more papers by this author
  • H. Medrano

    1. Laboratori de Fisiologia Vegetal, Grup de Biologia de les plantes en condicions mediterrànies, Universitat de les Illes Balears, Carretera de Valldemossa Km 7.5, 07122 Palma de Mallorca, Balears, Spain
    Search for more papers by this author

Author for correspondence: Jaume Flexas Tel: +34 971 172365 Fax: +34 971 173184 Email: jaume.flexas@uib.es

Summary

  • • Rubisco activity decreases under water stress, for reasons as yet unclear. Here, the covariation of stomatal conductance (gs) and relative water content (RWC), often observed during water stress, was impaired to assess the separate effects of these factors on Rubisco activity.
  • • Three different treatments were applied to soybean (Glycine max) and tobacco (Nicotiana tabacum): leaf desiccation (LD), in which stomatal closure was accompanied by large decreases of RWC; water stress (WS), in which minor decreases of RWC were observed along with stomatal closure; and exogenous application of abscisic acid (ABA), which triggered stomatal closure without changing RWC.
  • • Decreased RWC did not induce decreased initial Rubisco activity, which was impaired only in soybean by 40% when the gs dropped below 50 mmol m−2 s−1, regardless of the treatment. The mechanism for decreased activity differed among treatments, owing to decreased activation in LD and to total activity and protein content in WS and ABA.
  • • Despite the occurrence of Rubisco regulation, CO2 availability in the chloroplast, not impairment of Rubisco activity, limits photosynthesis during WS.

Introduction

Water stress is considered as the main environmental factor limiting plant growth and yield worldwide, especially in semi-arid areas (Boyer, 1982; Chaves et al., 2003). It is well known that one of the primary physiological consequences of drought is photosynthesis inhibition (Chaves, 1991; Cornic, 1994; Lawlor, 1995). There is now substantial consensus that reduced CO2 diffusion from the atmosphere to the site of carboxylation – as a result of both stomatal closure and reduced mesophyll conductance – is the main cause of decreased photosynthesis under most water stress (WS) conditions (Chaves & Oliveira, 2004; Flexas et al., 2004, 2006; Ennahli & Earl, 2005; Grassi & Magnani, 2005).

Nevertheless, WS has been shown to induce metabolic impairment and, particularly, decreases in the activity of Rubisco (Boyer, 1976; Lawlor, 1995; Flexas & Medrano, 2002; Lawlor & Cornic, 2002; Tezara et al., 2002; Bota et al., 2004). While decreased Rubisco activity may not be the cause of photosynthetic reduction during WS, its down-regulation may still be important because it could preclude a rapid recovery upon rewatering (Ennahli & Earl, 2005). The reported effects of WS on the activity of Rubisco differ among reports, from strong reductions of Rubisco activity per unit leaf area (Maroco et al., 2002) or per mg of protein (i.e. specific activity; Parry et al., 2002), to no effect (Lal et al., 1996; Pankovic et al., 1999; Delfine et al., 2001; Pelloux et al., 2001). These discrepancies probably arise from interspecific differences and/or differences in the WS intensities achieved.

It has been proposed that decreased Rubisco activity during WS is a direct consequence of the ionic conditions created inside cells as cell water content decreases (Gunasekera & Berkowitz, 1993; Lawlor & Cornic, 2002; Parry et al., 2002), which can be assessed by means of leaf relative water content (RWC). However, Bota et al. (2004) showed that the RWC at which Rubisco activity was decreased strongly depended on the species. On the other hand, Flexas & Medrano (2002) and Flexas et al. (2004) showed that Rubisco activity under WS responded to a certain threshold of daily maximum stomatal conductance (gs), which was common among C3 plants. Rubisco activity remains essentially unaffected by WS whenever gs is > 50–100 mmol m−2 s−1, regardless of the species. By contrast, when the gs value falls below that threshold, Rubisco activity may eventually decline, perhaps depending on the species and/or environmental conditions. This observation led to the hypothesis that stomatal closure triggers the down-regulation of Rubisco through its effects on chloroplast CO2 concentration (Flexas et al., 2006). Alternatively, stomatal closure and decreased Rubisco activity may share some common action [e.g. a response to abscisic acid (ABA)] that is triggered in response to WS.

Nevertheless, it is difficult to assess which of these two factors (i.e. decreased cell water content or stomatal closure) trigger the down-regulation of Rubisco activity, as gs and RWC often decline in parallel during water stress (Lawlor & Cornic, 2002; Bota et al., 2004; Ennahli & Earl, 2005). The aim of the present study was to assess separately the effects of both factors on Rubisco activity and related parameters by impairing the covariation of gs and RWC usually observed during water stress. This was achieved applying three different treatments to soybean and tobacco plants: leaf desiccation (LD); WS imposition; and exogenous application of abscisic acid (ABA).

Materials and Methods

Plant material and growth conditions

Soybean (Glycine max L.) and tobacco (Nicotiana tabaccum L.) were grown from seeds. Before sowing, soybean seeds were treated with 0.5% (v/v) NaOCl for 10 min and allowed to swell in distilled water for 2 h with continuous bubbling of air. Tobacco seeds were not pretreated. Seeds were sowed in separate trays of a mixture of sand and perlite (1 : 1) and placed in a growth chamber at a constant temperature (25°C) and a photon flux density of 600 µmol m−2 s−1 with a 14-h light:10-h dark regime. Seedlings were watered twice a day. Fifteen days after germination, plants were placed either in substrate or hydroponics solution. For plants growing in substrate, 4-l pots containing a mixture of perlite, horticultural substrate and clay were used.

Once the plants were established, the environmental conditions during growth were set to a 12-h photoperiod (25°C day: 20°C night), 50% relative humidity and a photon flux density, at plant height, of c. 800–1000 µmol m−2 s−1 (halogen lamps). Substrate plants were daily irrigated at pot capacity with 50% Hoagland's solution until the onset of the experiment, while for plants in hydroponics, solution was added several times per day in order to keep the entire root system submersed. All plants were 8–10-wk old during the experiment.

Treatments

Three different treatments were established: leaf desiccation (LD treatment); WS (WS treatment); and exogenous application of ABA (ABA treatment).

LD treatment was achieved by cutting the leaf petiole in air and letting the leaf desiccate for ≥ 1 h under ambient conditions. Leaves from both solid substrate- and hydroponic-grown plants were used. Several degrees of LD were developed during the time course of desiccation. The LD treatment was repeated in three to four leaves per species.

WS treatment was applied by stopping irrigation. Different degrees of WS were achieved over time. At least five different plants per species were subjected to WS.

ABA treatment consisted of a single application of 100 µm ABA (previously dissolved in 50% Hoagland's with 1 ml methanol) to the nutrient solution of plants grown in hydroponics. Also in this case different degrees of ABA action (as determined by gs) developed over time, as the applied ABA was progressively taken up by roots. Three to four different plants per species were subjected to this treatment.

Gas-exchange and chlorophyll fluorescence measurements

Leaf gas exchange parameters were measured simultaneously with measurements of chlorophyll fluorescence using an open gas exchange system (Li-6400; Li-Cor, Inc., Lincoln, NE, USA) with an integrated fluorescence chamber head (Li-6400-40 leaf chamber fluorometer; Li-Cor, Inc.). All measurements were carried out at 25°C and at 1500 µmol m−2 s−1, with 10% blue light. Such light intensity was previously determined to be just above light saturation for both species (data not shown). Cuvette CO2 concentration (Ca) was set at 400 µmol CO2 mol−1 air and vapour pressure deficit was kept at 2.0 ± 0.2 kPa.

In addition to net photosynthesis (AN) and gs, the incorporated fluorometer allowed determination of the actual photochemical efficiency of photosystem II (φPSII). This was determined by measuring steady-state fluorescence (F′) and maximum fluorescence during a light-saturating pulse of c. 8000 µmol m−2 s−1 (inline image), following the procedures of Genty et al. (1989):

image

The electron transport rate (ETR) was then calculated as:

ETR = φPSII · PFD · α

(α, a term which includes the product of leaf absorptance and the partitioning of absorbed quanta between photosystems I and II; PFD, photosynthetically active photon flux density).

α was previously determined for each treatment as the slope of the relationship between φPSII and inline image (i.e. the quantum efficiency of gross CO2 fixation) obtained by varying light intensity under nonphotorespiratory conditions in an atmosphere containing < 1% O2 (Valentini et al., 1995). α was found to be 0.51 in soybean and 0.33 in tobacco, with no difference observed between treatments.

From combined gas-exchange and chlorophyll fluorescence measurements, the chloroplast CO2 concentration (Cc) was calculated according to Epron et al. (1995). The model assumes that all the reducing power generated by the electron transport chain is used for photosynthesis and photorespiration, and that chlorophyll fluorescence gives a reliable estimate of the quantum yield of electron transport. Therefore, the ETR measured by chlorophyll fluorescence can be divided into two components:

ETR = ETRA + ETRP

(ETRA, the fraction of ETR used for CO2 assimilation; ETRP, the fraction of ETR used for photorespiration). ETRA and ETRP can be solved from data of AN, mitochondrial respiration in the dark (RD) and ETR, and from the known stochiometries of electron use in photosynthesis and photorespiration, as follows (Epron et al., 1995; Valentini et al., 1995):

ETRA = 1/3[ETR + 8(AN + RD)]

ETRP = 2/3[ETR − 4(AN + RD)]

From ETRA and ETRP, and the Rubisco specificity factor (τ), Cc can be calculated according to Laing et al. (1974), as follows:

τ = [(ETRA/ETRP)/(Cc/O)]

The oxygen molar fraction at the oxygenation site (O) was assumed to be equal to the molar fraction in the air, and a unique value of 85 was used in both species for τ (Parry et al., 1989). The main advantage of this method over others (e.g. Harley et al., 1992) is that estimation of Cc does not require previous knowledge about the substomatal CO2 concentration (Ci), which cannot be determined properly using conventional gas exchange systems when patchy stomata closure occurs, as could be the case at least for the ABA and detachment experiments (Terashima, 1992). Although chlorophyll fluorescence images obtained for tobacco (but not for soybean) using a FluorCAM (PSI Instruments, Brno, Czech Republic) suggested no incidence of patchiness in any of the treatments (data not shown), it would still be preferable to use a Ci-independent method.

Leaf water status

After gas exchange and chlorophyll fluorescence measurements, leaf discs from the same tobacco leaves were sampled. In the case of soybean plants, leaflets different from those used for gas exchange were sampled. Leaf RWC was determined in these leaf discs as follows:

RWC = [(fresh weight − dry weight) ÷ (turgid weight − dry weight)]

Turgid weight was determined by placing samples in distilled water and maintaining them at 4°C in darkness (to minimize respiration losses) until they reached a constant weight. Full turgor was typically reached after 12 h. Dry weight was obtained after placing the samples in an oven at 70°C for 48 h. Five to six replicates per species and treatment were obtained.

Biochemical determinations

For measurements of Rubisco activity and total soluble protein content, two discs (5.3 cm2) from the same leaves used for gas exchange measurements were freeze-clamped into liquid nitrogen and stored at −70°C before analysis.

The samples were ground to a fine powder in a mortar previously chilled with liquid nitrogen and were then homogenized in 1 ml of an ice-cold extraction medium. The extraction medium contained: 50 mm Bicine, pH 8; 20 mm MgCl2; 50 mmβ-mercaptoethanol; 2 mm phenylmethylsulfonyl fluoride; 30 mg of polyvinylpolypyrrolidone; 2 mm benzamidine; 2 mmɛ-amino-n-caproic acid; and 1% (v/v) protease inhibitor cocktail (Sigma Cat. P9599; Sigma, Dorset, UK). Extracts were clarified by centrifugation (11 000 g at 4°C for 2 min) and the supernatant was immediately assayed at 25°C for Rubisco activity. The initial and total Rubisco activities were determined according to Parry et al. (1997), and expressed on a per area basis. This method, like the typical spectrophotometric method (Ward & Keys, 1989), underestimates the actual activity of different species by 2.6-fold, on average. Nevertheless, the values may be qualitatively comparable (Rogers et al., 2001). For this reason, we will discuss the data, in terms of percentage variation, with respect to control plants. Total soluble protein was determined according to the method of Bradford (1976).

Statistical analysis

One-way analysis of variance (anova) was applied to assess the differences between treatments for each parameter and species. Differences among means were established using a Duncan test (P < 0.05). The data were analysed applying the SPSS 10.0 program for Windows (SPSS, Chicago, IL, USA).

Results

Comparison of control plants from soil and hydroponics

No significant differences between irrigated plants grown in soil and control plants grown in hydroponics solution were observed in gas exchange, protein content or Rubisco activity. In soybean, only leaf RWC differed between the two groups of plants, being significantly higher in plants grown in hydroponics (87.5 ± 1.0%) than in plants grown in solid substrate (75.1 ± 1.1%). Because the detachment experiments used leaves from both groups, the average RWC was an intermediate value between the two groups (80.1 ± 6.1%). In tobacco, there were no significant differences in RWC among the three groups (75.5 ± 2.3%).

When data for substrate-grown and hydroponic-grown control plants were combined, the initial Rubisco activity averaged 0.25 ± 0.02 µmol min−1 mg−1 protein in soybean and 0.22 ± 0.02 µmol min−1 mg−1 protein in tobacco. These values agree well with values previously reported for both species (e.g. Van Heerden et al., 2003 for soybean, Parry et al., 2002 for tobacco). The activation state averaged 83.3 ± 1.2% in soybean and 71.1 ± 4.1% in tobacco, also in agreement with previously published reports.

The relationship between gs and RWC

Similar ranges of gs were achieved by the three different treatments, albeit at different rates (Fig. 1). In detached soybean leaves (Fig. 1a), the gs dropped from c. 300 mmol m−2 s−1 to < 50 mmol m−2 s−1 in c. 1 h. In WS-treated plants, a similar decrease in gs was observed approx. 1 wk after withholding water (Fig. 1b). Finally, a similar decrease of gs in ABA-treated plants required c. 3 d (Fig. 1c). The aim of separating changes in gs from those in RWC was fully accomplished using these three treatments. Hence, in LD leaves, changes in RWC paralleled those in gs (Fig. 1a), and values as low as 55% were achieved c. 1 h after cutting the petiole in air. However, in WS leaves, and despite similar range of gs, the RWC was unaffected by the treatment during the first 6 d (Fig. 1b), decreasing only thereafter and to a lesser extent (70%) than in LD. Finally, in ABA-treated leaves, RWC was > 90% during the entire experiment (Fig. 1c), thus being totally independent of gs. Very similar results were obtained in tobacco (data not shown).

Figure 1.

Representative time courses of stomatal conductance (gs, closed symbols) and leaf relative water content (RWC, open symbols) in soybean (Glycine max) in response to three different treatments: (a) leaf desiccation (LD), consisting of cutting the leaf petiole in air; (b) water stress (WS), imposed by withholding water to potted plants; and (c) exogenous addition of 100 µm abscisic acid (ABA) to the medium in plants growing in hydroponics solution. Each treatment was imposed at time 0 (note the different timescales among treatments).

Rubisco activity as related to RWC and gs

No clear relationship was found between initial Rubisco activity and RWC when all data were pooled together (Fig. 2). However, when values for the different treatments were grouped by gs, a pattern emerged in soybean, which was independent of the treatment (Fig. 3a). Basically, the initial Rubisco activity remained unaffected by any treatment until the gs dropped below 50 mmol m−2 s−1. Below this threshold, the initial Rubisco activity was decreased by 30–40%, regardless of the treatment. By contrast, the Rubisco activity in tobacco seemed to be more resistant, dropping at the same gs threshold only in ABA plants and to a lesser extent (20%) than in soybean (Fig. 3b).

Figure 2.

The relationship between initial Rubisco activity on an area basis (% of control values) and leaf relative water content (RWC) in (a) soybean (Glycine max) and (b) tobacco (Nicotiana tabacum). Data are single values for each treatment: exogenous addition of abscisic acid (ABA, black symbols), water stress (WS, white symbols) and leaf desiccation (LD, grey symbols).

Figure 3.

Initial Rubisco activity on an area basis (% of control values) at three different intervals of stomatal conductance in (a) soybean (Glycine max) and (b) tobacco (Nicotiana tabacum). Three different treatments were applied: exogenous addition of abscisic acid (ABA, black bars), water stress (WS, light grey bars) and leaf desiccation (LD, dark grey bars). Values are means ± standard error of four to eight replicates. An asterisk indicates a significant difference (P < 0.05) between the treatment and its control (i.e. the same treatment at gs > 150 mmol m−2 s−1).

By contrast, other parameters related to Rubisco activity, such as total soluble protein content, total Rubisco activity or Rubisco activation state, did not follow a general pattern independent of the treatment. For instance, the total soluble protein content in soybean decreased significantly when the gs dropped below 50 mmol m−2 s−1 in WS- and ABA-treated plants, but not in LD-treated plants (Fig. 4a). In tobacco, however, there was no effect of ABA on total soluble protein, and treatment with WS resulted in a progressive increase (Fig. 4b), so that total soluble protein at the end of the experiment was up to 40% higher than in controls. In soybean, as for total soluble protein, total Rubisco activity declined at the lowest gs range only in WS- and ABA-treated plants (Fig. 5a), but to very different extents (50% in WS as compared to 20% in ABA). The total Rubisco activity declined only by 20% in ABA-treated tobacco plants (Fig. 5b). Finally, the Rubisco activation state declined at the lowest gs range only in LD-treated soybean plants (Fig. 6a), but not in tobacco (Fig. 6b).

Figure 4.

Total soluble protein content on an area basis (% of control values) at three different intervals of stomatal conductance in (a) soybean (Glycine max) and (b) tobacco (Nicotiana tabacum). Three different treatments were applied: exogenous addition of abscisic acid (ABA, black bars), water stress (WS, light grey bars) and leaf desiccation (LD, dark grey bars). Values are means ± standard error of four to eight replicates. An asterisk indicates a significant difference (P < 0.05) between the treatment and its control (i.e. the same treatment at gs > 150 mmol m−2 s−1).

Figure 5.

Total Rubisco activity on an area basis (% of control values) at three different intervals of stomatal conductance in (a) soybean (Glycine max) and (b) tobacco (Nicotiana tabacum). Three different treatments were applied: exogenous addition of abscisic acid (ABA, black bars), water stress (WS, light grey bars) and leaf desiccation (LD, dark grey bars). Values are means ± standard error of four to eight replicates. An asterisk indicates a significant difference (P < 0.05) between the treatment and its control (i.e. the same treatment at gs > 150 mmol m−2 s−1).

Figure 6.

Rubisco activation state (%) at three different intervals of stomatal conductance in (a) soybean (Glycine max) and (b) tobacco (Nicotiana tabacum). Three different treatments were applied: exogenous addition of abscisic acid (ABA, black bars), water stress (WS, light grey bars) and leaf desiccation (LD, dark grey bars). Values are means ± standard error of four to eight replicates. An asterisk indicates a significant difference (P < 0.05) between the treatment and its control (i.e. same treatment at gs > 150 mmol m−2 s−1).

Photosynthetic parameters in relation to gs and Cc

Pooling all data together, good correspondence was found between AN and gs (Fig. 7). In addition, good correspondence was found between gs and Cc. As stomata closed in response to any of the applied treatments, the Cc decreased from c. 250 µmol mol−1 to c. 50 µmol mol−1, although the exact pattern differed slightly between species (Fig. 8). As a result of these relationships, the AN was strongly correlated to Cc, regardless of the treatment, although the relationship was different in each species (Fig. 8). In soybean (Fig. 8a), a linear relationship was observed, with a maximum AN of c. 35 µmol of CO2 m−2 s−1 at the highest Cc, while in tobacco the relationship was somewhat curvilinear (Fig. 8b), with a maximum AN of only 25 µmol of CO2 m−2 s−1. Remarkably, in both species, the AN was close to zero when the Cc was c. 50 µmol mol−1 (i.e. close to the known CO2 compensation point for C3 plants), supporting the validity of Cc estimates.

Figure 7.

The relationship between net CO2 assimilation and stomatal conductance (gs) in (a) soybean (Glycine max) and (b) tobacco (Nicotiana tabacum). Data are single values for each treatment: exogenous addition of abscisic acid (ABA, black symbols), water stress (WS, white symbols) and leaf desiccation (LD, grey symbols).

Figure 8.

The relationship between chloroplast CO2 concentration (Cc) and stomatal conductance (gs) in (a) soybean (Glycine max) and (b) tobacco (Nicotiana tabacum). Data are single values for each treatment: exogenous addition of abscisic acid (ABA, black symbols), water stress (WS, white symbols) and leaf desiccation (LD, grey symbols).

In soybean, initial (Fig. 9a) and total (Fig. 9b) Rubisco activities did not show any clear dependency on Cc, except that they were never depressed at high Cc. By contrast, the Rubisco activation state showed a tendency to decline when the Cc dropped below 100 µmol mol−1 (Fig. 9c), although the number of samples may not be sufficient to confirm this tendency. In tobacco, no tendency was observed in the response of any of the parameters to Cc (not shown).

Figure 9.

The relationship between net CO2 assimilation and chloroplast CO2 concentration in (a) soybean (Glycine max) and (b) tobacco (Nicotiana tabacum). Data are single values for each treatment: exogenous addition of abscisic acid (ABA, black symbols), water stress (WS, white symbols) and leaf desiccation (LD, grey symbols).

Discussion

In the present experiment, three different treatments were applied: LD, WS and ABA. These three treatments resulted in (i) induction of almost complete stomatal closure at very different rates, from minutes (LD) to days (ABA) or up to a week (WS), and (ii) breaking the often observed covariation of gs and RWC during WS, with both parameters changing totally in parallel (LD), weakly related (WS) or totally unrelated (ABA).

Contrary to previous assumptions (e.g. Lawlor & Cornic, 2002), the present data clearly show that WS-induced decreases in Rubisco activity are not related to decreased RWC. For instance, an identical degree of inhibition was observed in soybean leaves desiccated to an RWC of 55% within 1 h and in leaves when stomatal closure was induced by ABA addition during 3 d while keeping an RWC of 90%. However, a gs threshold for decreased initial Rubisco activity is clearly shown, regardless of the treatment used (i.e. independent of the velocity of stomatal closure imposition as well as of RWC), thus supporting the view by Flexas et al. (2004, 2006) that gs is a good reference parameter for the down-regulation of Rubisco activity during WS. The present results also confirm that the down-regulation of Rubisco activity induced by stomatal closure is species-dependent, with soybean acting as a sensitive species and with tobacco acting as a resistant species.

However, the mechanism for decreased initial Rubisco activity seems to depend on the treatment applied. In principle, decreased initial Rubisco activity could be the consequence of decreased protein concentration, increased concentration of binding inhibitors (i.e. a decrease of the maximum activity), decreased activation state, or combinations of two or more of these factors (Parry et al., 2002; Bota et al., 2004). While in the present study we did not determine Rubisco content or maximum activity, measurements of total soluble protein content, total Rubisco activity and the Rubisco activation state may help to elucidate the mechanism responsible for decreased initial activity in each case. In soybean, the entire decrease of initial Rubisco induced by LD seems to be a result of the decreased activation state of the enzyme, because no decreases of either total soluble protein or total Rubisco activity were observed. By contrast, in WS- and ABA-treated plants, no significant reduction of the activation state was observed, but both total soluble protein and total Rubisco activity decreased. Therefore, in the case of WS- and ABA-treated plants, a decrease in the amount of Rubisco (and perhaps an increase of inhibitors) may be the reason for decreased initial activity. In tobacco, initial activity decreased only in ABA-treated plants, which was related to a decrease of total activity. A decrease in the content of Rubisco cannot be discounted, although in this case it cannot be assessed by means of changes in total soluble protein, as these may be masked by an increase in some other soluble protein, as observed in WS plants, probably an osmolyte (Kawaguchi et al., 2003).

Surprisingly, despite the described mechanistic differences, the reduction of initial activity was similar among treatments, suggesting a fine down-regulation in which several different mechanisms operate at different timescales. For instance, in response to almost complete stomatal closure within 1 h, as occurs with LD, an adjustment in the activation state – which is a biochemical response and therefore does not involve changes in the transcription and translation of Rubisco subunits – may be the only possible response. By contrast, when similar stomatal closure develops over several days, other mechanisms involving gene expression may operate, resulting in decreased amounts of Rubisco and/or an increased content of Rubisco inhibitors. In tobacco, for instance, both a decrease in the levels of mRNA encoding the small subunit of Rubisco (Kawaguchi et al., 2003), a decreased amount of Rubisco and increased levels of the ‘daytime inhibitor’ (Parry et al., 2002) have been observed in response to WS conditions similar to those used in the present study. As both ABA and WS may have induced early senescence (Olsson, 1995), this could be another reason for decreased Rubisco amount and activity, as shown in soybean (Jiang et al., 1993). Whatever the exact sequence of operating mechanisms, the physiological meaning of these adjustments is unknown and may deserve detailed attention in future studies.

Because gs, mesophyll conductance and Cc are coregulated during drought, it has been hypothesized that Cc, and not gs itself, may act as regulator of Rubisco content and/or activity, hence explaining the strong correlation with gs (Flexas et al., 2006). In the present study, net photosynthesis strongly correlated with Cc, as observed in WS experiments (Flexas et al., 2002). Remarkably, net photosynthesis approached zero when Cc was close to 50 µmol mol−1 (i.e. close to the CO2 compensation point), as already shown in water-stressed grapevines (Flexas et al., 2002). Altogether, these observations suggest that CO2 availability – and not decreased Rubisco activity – was the main cause for decreased photosynthesis, regardless of the treatment applied, as usually observed in WS (Flexas et al., 2004; Grassi & Magnani, 2005) as well as in ABA and LD experiments (Meyer & Genty, 1998). However, the correlation of Rubisco parameters with Cc was less clear (Fig. 7). If any, a Cc threshold of c. 100 µmol mol−1 was observed above which no changes in Rubisco parameters occur (Fig. 10). However, no clear tendency was observed below that threshold except, perhaps, for the activation state. Remarkably, a very similar CO2 concentration threshold for decreased activation of Rubisco was shown in studies in vitro (Perchorowicz & Jensen, 1983). Nevertheless, the lack of changes in tobacco at a similar Cc precludes any clear conclusion regarding the role of Cc in the regulation of Rubisco under WS.

Figure 10.

The relationship of (a) initial Rubisco activity, (b) total Rubisco activity and (c) Rubisco activation state and chloroplast CO2 concentration in soybean (Glycine max). Data are grouped using the same gs intervals as in Figs 2–5, and values are means ± standard error of four to eight replicates. Three treatments were applied: exogenous addition of abscisic acid (ABA, black symbols), water stress (WS, white symbols) and leaf desiccation (LD, grey symbols).

In summary, the present results clearly show that decreased initial Rubisco activity during WS is not mediated by decreased RWC, but somehow related to gs, although it is not yet clear whether stomatal closure acts on Rubisco regulation through its effects on Cc. Bota et al. (2004) had shown previously that the mechanism for decreased Rubisco activity (i.e. decreased amounts of Rubisco vs. decreased activation state) differed among species. The present results show that the mechanism may differ even within a single species, depending on how stomatal closure is induced and/or on the velocity of stomatal closure. However, decreased Rubisco activity was probably not the cause of photosynthesis depression in any of the applied treatments, and thus the physiological meaning of these fine adjustments of Rubisco activity in response to stomatal closure remains unknown and may deserve more detailed attention in future studies.

Acknowledgements

This work was supported by CICYT Projects BFI2002-00772 and BFU2005-03102/BFI (Plan Nacional, Spain). M. R.-C. was the beneficiary of a contract from Programa Ramón y Cajal (M.E.C.).

Ancillary